Insights

Claude Fable 5: we tested the model that got itself banned

Claude Opus 4.8: a sharp, sensible writer that just misses the top spot

Is AI overwriting women at work?

Gemini 3.5 Flash: Google’s newest model finishes last on our leaderboard

Claude Opus 4.7: a sturdy all-rounder that occasionally tries too hard

Sloppy videos aren’t an AI problem. It’s a you problem

The 3,000-year-old piece of clay that holds the secret to getting the best out of AI

GPT-5.5: strong on clarity, shaky on…everything else.

How to use PR for AI discoverability: the essential guide

The best AI image models

Becoming more than the sum of your prompts

AI in PR: How to raise standards instead of lowering them

Gemini 3.1 Pro: Google’s flagship just topped our leaderboard

© 2025 Definition. All rights reserved.